Major Insurers Drop AI Coverage as Startups Seize the Opportunity
Berkshire Hathaway, Chubb and Travelers have won state regulatory approval to exclude AI-related damages from standard corporate liability policies. State regulators approved more than 80% of these exclusion requests, with Florida, Connecticut and Maryland leading the approvals. The exclusions took effect as early as January 2026.
The move signals a hard truth: major carriers say they cannot price the risk. Until they understand what generative AI and LLM failures will cost, they are refusing to cover them.
What's Now Excluded
The Insurance Services Office, which sets industry standards, introduced two new optional endorsements. Carriers including Berkley have adopted absolute AI exclusions across directors and officers, errors and omissions, and fiduciary liability policies.
These endorsements remove coverage for bodily injury, property damage and personal or advertising injury tied to AI outputs. That includes defamatory content, intellectual property infringement and physical damages from AI-driven errors.
The reach is broad. ISO forms underpin roughly 82% of U.S. property and casualty policies. Adoption is expected to accelerate at renewal cycles.
Businesses now face exposure across multiple claim categories: copyright and IP claims from large language model training, privacy and data-use claims, antitrust allegations, discrimination claims tied to algorithmic bias and AI-related securities lawsuits. Courts have not yet settled how broadly these exclusions apply, leaving some policyholders with coverage that may prove illusory.
Small to mid-sized firms without specialized coverage face the greatest exposure.
A Market Takes Shape
Specialized insurers are filling the gap. Munich Re and startups including Corgi, Armilla, Mayflower Specialty and Embroker now offer standalone AI liability policies.
Coverage limits range from $2 million to $50 million. Annual premiums span from a few hundred dollars to several hundred thousand dollars, depending on risk profile and coverage limits.
This mirrors what happened with cybersecurity a decade ago. A wave of attacks triggered corporate claims. Businesses argued traditional policies covered the losses because they did not explicitly exclude them. Insurers carved cyber out of standard coverage and built standalone products. That market matured. The AI liability market is following the same path.
The Industry's Own Contradiction
Insurance giants are deploying AI for insurance agents to orchestrate workflows across claims, underwriting and policy servicing. Those same carriers are simultaneously pulling AI coverage from the policies they sell.
Insurers are betting on AI internally while refusing to absorb AI risk externally. They are not declaring AI uninsurable. They are saying they do not yet know what it costs.
What Businesses Should Do Now
Review existing policies with your broker for new AI exclusion endorsements. The endorsements are likely to appear at renewal if they have not already.
Consider three options:
- Affirmative AI coverage through technology errors and omissions policies
- Cyber liability insurance with AI components
- Emerging standalone AI liability products from specialized carriers
Strengthen internal AI governance. Conduct bias testing on AI systems. Disclose AI use in business processes. These steps reduce exposure and may lower premiums.
Until insurers develop pricing models for AI risk, the cost of that risk sits with the businesses deploying the technology.
Your membership also unlocks:
